A data‐driven maintenance policy for railway wheelset based on survival analysis and Markov decision process

Abstract

Wheelsets absorb a significant part of the maintenance budget of any train operating company. Although wheel wear has been an extensively discussed topic in the literature, wear rates are very rarely characterized by using degradation data in a real‐world case study aimed at identifying optimal maintenance policies including both degradation and recovery modeling. Furthermore, wheel defects, which impose an additional challenge to the modeling of the lifecycle of the wheels, are usually considered separately in the literature. In this study, conducted at a Portuguese train operating company, 17 years of inspection data are used to estimate wheel wear rates and survival curves, which are further incorporated into a Markov decision process (MDP) model. A bidimensional framework considering discrete intervals of wheel diameter along with a quantitative variable (kilometers since last turning/renewal) is used to represent the possible wheel states, while the probability of a defect interfering with the wheel maintenance schedule is modeled by contemplating survival curves derived from a Cox proportional‐hazards model. Optimal results in terms of minimal cost policy are discussed in the context of the MDP, but a more realistic and easy‐to‐implement policy fixing one of the parameters is compared with the optimal policy. Results showed that in practice train operating companies might benefit from using the easy‐to‐implement policy, which has an associated long‐run average cost only about 1% higher than the one suggested by the optimal decision map.

Strategical selection of maintenance type under different conditions

“…By examining failure rates, causes of failure, and the time between failures, maintenance professionals can make informed decisions about preventive or corrective actions 18 . Statistical methods such as failure distribution analysis 19 , reliability modeling [19][20][21] , and survival analysis 22,23 are commonly used for this purpose.…”

Using expert judgement techniques to assess reliability for long service-life components: An application to railway wheelsets

“…For example, regarding the reliability of wheelsets, real degradation data obtained from inspection activities should be used 41 to compare with the results obtained from the proposed method and to support models that estimate real wear rates and damage occurrences, and provide optimal maintenance and inspection intervals to reduce the lifecycle cost in the long-term of wheelset components. 42,43 Such results and reliability curves derived from expert judgement techniques should also be integrated with maintenance planning and maintenance scheduling models, in order to improve the assets availability, reliability and the associated operational costs. 44,45…”

Reliability‐based fault analysis models with industrial applications: A systematic literature review

“…Data are one of the key elements in ML to effectively apply smart algorithms to detect faults to enhance asset reliability in the industry. Internet of things (IoT) provide a platform to collect real‐time data, which is a basic requirement in FDD models for a wide variety of industrial applications (Kanawaday and Sane, 2017; 202 Civerchia et al 2017; 203 de Almeida Costa et al, 2020 204 ). IoT capabilities along with other technologies such as cloud computing enables to generate, store, and clean data to be used in applications such as FDD.…”